The rediscovery of Camponotus (Myrmaphaenus) yogi Wheeler (Hymenoptera: Formicidae).

Reprinted from Psyche, Vol. 73, No. 3, 1966

THE REDISCOVERY OF

CAMPONOTUS (MYRMABHAENUS) YOGI WHEELER

(HYMENOPTERA: FORMICIDAE)

By Wm. S. Creighton^ City College, New York'^
and Roy R. Smelling, Los Angeles County Museum^

No North American ant has been more of an enigma than
Wheeler's Camponotus yogi. Much of this is due to the fact that the
two types of yogi were misplaced when a part of the Wheeler Col-
lection was transferred from the Museum of Comparative Zoology to
the American A'luseum of Natural History in 1938. For the next
twenty-six years Wheeler's description of yogi was the only source of
information on this species (i). Of necessity this description was
the basis for the treatment of yogi in The Ants of North America
(2). Unfortunately, one of the key features chosen (the length of
the antennal scape) was incorrectly described by Wheeler; hence
the key for yogi in the above publication is confusing rather than
helpful. It appears that yogi has been saved from even greater con-
fusion only because so little additional material has been taken. In
1958 F. Raney found a few specimens at the Oak Creek Ranger
Station in San Diego County, California. In 1963 the junior author
took three colonies of yogi at Etiwanda, in San Bernardino County,
California. The senior author at first refused to believe that this
material could be yogi because it so obviously failed to agree with
Wheeler's description. Then, in 1964, the two types of yogi were
discovered in the collection of the American Museum of Natural
History. These established the fact that the Oak Creek and Etiwanda
specimens are yogi.

Several of the shortcomings of Wheeler's description can be at-
tributed to his attempt to relate yogi to the subgenus Colobopsis.
Whether he realized it or not, Wheeler described important features
of the head of the major of yoffi from a position where it most
closely resembled that of a Colobopsis major. That is to say, the
head was not viewed in full face but tilted forward until the trun-
cated anterior portion and the mandibles were barely visible. There
is no possible doubt about this for the "broadly excised posterior
border" which Wheeler described for the head of the yogi major

"Emeritus Professor, Department of Biology
"Senior Preparator, Department of Entomology

Published with a Grant-in-Aid of Research from the Society of the Sigma
Xi. Manuscript received by the editor May 9, 1966

1 88 Psyche [September

cannot be seen unless the head is tilted forward. In conventional
full face view the occipital border of the major of yogi is flat or very
slightly convex (See Plate 12, Fig. 2). Precisely the same explanation
applies to Wheeler's statement that the scapes of the major of yogi
"reach the posterior corners of the head". When the head of the
major is tilted far enough forward to show the excavated rear border
of the occiput, it is true that the scapes appear to reach the occipital
corners. But in conventional full face view they extend beyond the
occipital corners. It is obvious that tilting the head would not greatly
affect the configuration of the frontal carinae and, if one is willing
to unravel the description that both Wheeler and Forel customarily
applied to lyrate carinae, then these were accurately described. In the
writer's opinion this is the only feature in the description of yogi
which enabled Emery to allocate the species to the subgenus Myrma-
phaenus for, with strongly lyrate frontal carinae, it cannot belong
to Colohopsis.

Finally, there is the matter of pilosity. Wheeler recognized that
some of the cephalic pilosity of the major of yogi is notably different
from that on other parts of the body. He gave an accurate descrip-
tion of the short, blunt, reclinate hairs that arise from the foveolae
of the clypeus and the adjacent parts of the genae. But he failed to
mention the erect hairs which occur on the sides of the head from
the rear border of the eye to the mandibular insertion. These hairs
also occur on the lateral parts of the gula and over an area extending
inward toward the antennal fossa. Although short, they are quite
numerous and form a conspicuous fringe on the anterior half of the
head. Their structure is unusual for, although each hair is of uni-
form thickness throughout most of its length, many of them have
enlarged tips. The enlarged tip is often spherical and such hairs look
remarkably like the upper ends of insect pins.

One of the most characteristic features of yogi is the sculpture on
the front of the head of the major. The surface is rough and covered
with coarse elevations and depressions which are too broad and ill-
defined to be called rugae. The roughened areas on the clypeus
have no fixed direction but those on the genae approximately parallel
the long axis of the head. Among these roughened ridges are scat-
tered oval foveolae from which flattened hairs arise, but these are
difficult to see because the entire roughened surface is evenly covered
with very fine, densely set, oval punctures which give the area a

Explanation of Plate 12
Camponotus (Myrmaphaenus) yogi Wheeler. Figure 1. Head of female.
Figure 2. Head of major. Both figures drawn to the same scale.

190 Psyche [September

granular appearance. The sculpture just described is confined to the
anterior half of the head. The frontal lobes are heavily shagreened,
the middle of the occiput more feebly shagreened and the sides of
the head and the occipital angles are smooth and shining. There
are conspicuous circular punctures on the frontal lobes, smaller oval
ones on the sides of the head and very small and obscure punctures on
the occiput. No other North American species of Camponotus has
a comparable cephalic sculpture and yogi may be recognized by this
feature alone.

There is nothing in Wheeler's treatment of yogi more exasperat-
ing than his statement that this species is related to Colobopsis. In
i8g6 Emery listed the distinguishing features of Colobopsis (3).
When Wheeler monographed our North American forms in 1904
he repeated Emery's criteria and added two of his own (4). Ac-
cording to Wheeler's summary the female and major worker of Colo-
bopsis have a head in which the truncated portion is circular in outline
and sharply separated from the remainder of the head (marginate).
There are conspicuous, umbilicate punctures on the sides of the head
immediately behind the truncation. The mandibles have an external
ridge or angle which separates the anterior face from the equally
large latero-ventral face. Medias are rare or lacking and the pupae
are not enclosed in cocoons. Because of the limited type material of
yogi Wheeler could not know that in this species medias are present
and the larvae are enclosed in cocoons. Nor could he know anything
about the structure of the female. Nevertheless, Wheeler was aware
that the truncated portion of the head of the major of yogi is neither
circular in outline nor marginate. He was aware that there is no
external ridge on the major's mandible. Since he described the dense,
granular, sculpture which obscures the foveolae on the sides of the
head of the major of yogi, Wheeler must have realized that these
foveolae are scarcely comparable to the distinct, umbilicate, punctures
of the Colobopsis major. In short, not a single feature of the major
of yogi agreed with the m.ajor of Colobopsis as that caste was defined
by Wheeler in 1904.

These contradictions are annoying but they are not inexplicable.
Wheeler's initial views on Colobopsis were based almost entirely on
species in the truncatus-impressus complex, as were the views which
Emery had expressed earlier. By 1907, due to the many identified
exotics which he had received from Forel, Wheeler was prepared to
expand his original views on Colobopsis. It is regrettable that when
he described yogi Wheeler failed to make it clear that some of the
Asiatic and South Pacific species assigned to Colobopsis are more

1966] Creighton and Snelling ~ Camponotus 191

aberrant than yogi when compared to species in the truncatus-im-
■pressus complex. With this in mind the question is not whether
Wheeler was justified in placing yogi in Colobopsis but whether
i.mery was any better off when he transferred it to Myrmaphaenus.
At present no final answer can be given for both subgenera are un-
usually heterogeneous. Nevertheless we prefer Emery's treatment for
the following reasons:

(i ) It IS quite impossible to relate yogi to the truncatus-impresms
complex in Colobopsis.

(2) If yogi is assigned to Colobopsis it will have to be placed in
one of the Old World groups which are, at present, too ill-defined to
permit certainty of assignment.

(3) There is no feature of yogi which would prevent its inclusion
in Myrmaphaenus and it possesses several features which indicate that
It fits better in that subgenus than in Colobopsis.

Most of these features have been mentioned above but one of them
merits a more detailed discussion. The head of the female of yogi
resembles that of the media rather than that of the major. In full
face view the head of the major is as wide (in some specimens slightly
wider) at the level of the rear of the clypeus as it is at the anterior
border of the eyes. This gives the head a distinctly rectangular out-
line, for the sides turn in toward the mandibular insertions abruptly
below the level of the middle of the clypeus. In the female and
media the sides of the head converge gradually from the level of the
anterior border of the eyes to the mandibular insertions. Thus in
full face view the anterior half of the head is distinctly narrower
than the posterior half (See Plate 12, Fig. i). There are sculptural
differences as well, for the female and the media lack the roughened
areas on the clypeus and its surface, as well as that of the truncated
parts which flank it, is granulo-shagreened only. The roughened
ridges are present on the sides of the head but they are feebler than
those of the major and the oval foveolae show more plainly.

A comparable condition is found in C. (Myrmaphaenus) andrei
i^orel, Which occurs on the Mexican plateau. Moreover, the major
of andret is remarkably like that of yogi, both in the anterior trun-
cation of the head and in the structure of the clypeus.^' Since andrei
has a strongly polymorphic worker caste it would appear that we are
reaching some uniformity in the species assigned to Myrmaphaenus.
At least we have in yogi, andrei and ulcerosus three species \vhere

""The insect which Forel described as the major of andrei is actually a
m«d,a Neither he nor Wheeler ever saw the major of andrei, which is a
very distmct caste.

192

Psyche [September

the worker caste is polymorphic, the front of the head of the major
is truncated, the clypeus of the major is flat and ecarinate or nearly
so, the anterior border of the clypeus of the major is distinctly (often
deeply) impressed and the head of the female resembles that of the
media. These three species form a reasonably compact geographic
group, since yogi occurs in southern California, ulcerosus m western
Texas, southern Arizona and northern Chihuahua and andrei m
Durango, Zacatecas and Hidalgo. It seems best to recognize the
common structural and geographical affinities of these species and we
believe that Emery's assignment of yoffi to the subgenus Myrma-
phaenus is the correct procedure.

The meager biological data which we can present at this time are
as tantalizing as the taxonomic history of yogi has been confusing.
The type specimens were taken, according to Wheeler, "from a
hollow twig of manzanita ... on Point Loma, near San Diego."
The senior author has examined considerable manzanita and live oak
in coastal southern California without successfully relocating the
species. Quite by accident the junior author was led to an examina-
tion of the living stems of Haplopappus pinifolius (Gray) Hall.

This shrub occurs in the foothill chaparral of the San Gabriel,
San Bernardino and San Jacinto Mountains south to San Diego
County. Although this species does not occur on Point Loma (the
type locality of yogi), a related species, H. palmeri (Gray) Hall, is
found there. It is possibly significant that all the records of yogi
thus far are from areas where one or the other of these species of
Haplopappus is known to occur.

To date, the junior author has taken six colonies of yogi, all from
stems of H. pinifolius. The most unusual point of this association
is that all six colonies were taken from living stems. Numerous
dead stems were examined, but only traces of former occupancy by
yogi were found in these. So far as is known the ant does no excava-
tion of the living tissues; all galleries are merely appropriated bur-
rowings of buprestid larvae, apparently of one or two species of
A cmaeodera. The manner in which the founding queen initially
gains entry into the beetle burrows is not known.

Attempts to discover something of the foraging activities of yogi
have so far been futile. No foraging individuals have been seen
during the day or at night. However, observations on a captive

Explanation of Plate 13

Camponotus (Myrmaphaenus) yogi Wheeler. Fig. 1. Male. Fig. 2. Minor.
Fig. 3. Major. Fig. 4. Female. All figures drawn to the same scale.

194 Psyche [September

colony suggest that this species regularly forages at night. During
the day, members of the captive colony normally remained crowded
within galleries cut into a wood block; with the coming of darkness
they moved out of the tunnels to the surface of the block to feed
on a honey-water solution provided for them.

The captive colony showed no interest in dead insects nor in
various vegetable foods (seeds, bran, wheat germ, pollen, etc.). The
primary food source seems to be the sweet exudates of several species
of Pseudococcidae. Species of mealybugs are: Puto sp.*, (immatures
only), Humococcus inor?iatus McKenzie'^, Anisococcus crawii
(Coquillett)" and Chorizococcus abroniae McKenzie?". The first
two species were taken from the galleries occupied by the ant, and
were feeding on the plant within these galleries. In the case of the
two latter species there is some confusion, since both were mixed in
a single vial ; some of the specimens are known to have come from
the galleries while others were taken feeding on the outer surface of
the stems and crowns. Whether or not the ants actually moved the
mealybugs into the galleries remains to be determined. However,
that ants do transport mealybugs to advantageous feeding sites is
well known, and it would not be amiss to suggest that yogi does so.

Two ant species have been found living in a plesiobiotic relationship
with yog'i. There are Solenopsis (D.) molesta validiuscula Emery
and Leptothorax andrei Emery. Each of these was found only once,
and little can be said here regarding their relationship with yogi.
The former species is commonly associated with larger ants, and the
latter has been taken several times from colonies of various species
of Camponotus and Formica.

One colony was apparently being attacked at the time it was
collected. The attacker was Formica (F.) pilicornis Emery, a com-
mon and aggressive species in the lower chaparral region. The
Formica had gained entrance into the galleries of the Camponotus,
and workers of the former were seen carrying off the larvae and
pupae of yogi. The remainder of the yogi colony, soldiers, workers,
the queen and some brood, were crammed into three upper galleries,
the majors lowest and receiving the brunt of the attack. The majors
of yogi were massed together, with antennae extended forward ; when
a pilicornis worker touched the antennae the yogi major lurched
forward and attempted to fasten its jaws to an appendage of the
attacker. Several such encounters were noted: when the defender's

Metermined by H. L. McKenzie
Metermined by R. F. Wilkey

'^96^1 Creighton and Snelling — Camponotus 195

jaws closed over the appendage of a pilicornis, the yogi major jerked
back into the original position; at the same time the head was
snapped up and down. This frequently resulted in severing or severe-
ly mangling the appendage of the attacker.

It was noted that the Formica so attacked usually retreated in a
liighly agitated manner. Closer observation of the process revealed
that when the jaws of the yogi major clamped over an attacker's
appendage a grayish-white exudate oozed forth from the lower part
of the major's head. Creighton (5) noted a similar occurrence in
the case of C. (Colobopsis) papago, with the remark that the fluid
came from the mouth. Subsequent examination of the yogi majors
suggested that in the case of this species, it is exuded from the man-
dibular bases; individuals which had the hardened material quite
conspicuous on the mandibles and lower portions of the face showed
no traces of it in the immediate vicinity of the mouthparts. This
fluid, when in contact with the integument of ants of other species,
causes considerable excitation of the individual affected, and, in the
case of the pilicornis, an immediate retreat from the conflict. That
this fluid serves as a defensive repellent seems obvious; whether or
not it IS toxic as well is not yet known.

The colonies of yogi thus far collected have been rather small; it
is doubtful if colonies exceed three hundred individuals. Colony
no. I, collected on May 11, contained 27 majors and 165 media and
minor workers; the queen was not located, but is presumed to have
been overlooked during the collecting. Colony no. 4, collected on
September 29, contained i queen, 8 males, 54 majors, 90 media and
minor workers, 51 larvae and 7 pupae (5 cf cf, 2 ??) ; colony no. 5,
also taken on September 29, consisted of i queen, i alate female,
3 males, 12 majors and 36 media and minor workers, and is the
smallest colony taken to date.

The mating flight evidently takes place during late summer.
Winged sexuals have been taken in the nests in late August and
September; winged females have been attracted to lights at night in
Glendale, Los Angeles County, on October 14.

Literature Cited

1. Wheeler, W. M., Bull. Amer. Mus. Nat. Hist., 34: 420 (1915)

2. Creighton, W. S., Bull. Mus. Comp. Zool., 104: 401 (1950)

3. Emery, Carlo, Mem. Acad. Sci. Bologna, 5: 761-780 (1896)

4. Wheeler, W. M., Bull. Amer. Mus. Nat. Hist., 20: 140 (1904)

5. Creighton, W. S., Psyche, 59, 4: 161 (1952)

rt-vcijc, \'}f.'i

■\'ru. 7:-., l*T..\Tr. i:

rRFreriTO^' .^^n yniij.i^r. — fJhi.K^jjucT.-j

Psyche, 1966

Vol. 73, Plate 13

Creiohton and Snelling — Camponotus